[PATCH 1/1] staging: usbip: fix wrong endian conversion

[Arjan Mels]

Greg,

David's correction is indeed correct. My original code is the same as David 
proposes, something must have gone wrong during my patch creation process. 
(I do not really understand how, maybe wrong copy/paste, or maybe something 
weird happened with diff or line offsets.)

I did not catch it on my Intel box, because I did not use a webcam/dvb-t 
tuner on the intel box for testing only on my mips router (which is already 
big-endian, so the conversions do nothing).

Apologies for the extra work. David, thanks for catching this.

Best Regards,

Arjan


-----Original Message----- 
From: Greg KH
Sent: Thursday, May 12, 2011 17:05
To: David Chang
Cc: devel at driverdev.osuosl.org ; Arjan Mels
Subject: Re: [PATCH 1/1] staging: usbip: fix wrong endian conversion

On Thu, May 12, 2011 at 06:31:11PM +0800, David Chang wrote:
> Fix number_of_packets wrong endian conversion in function
> correct_endian_ret_submit()
>
> Signed-off-by: David Chang <dchang at novell.com>
> ---
>  drivers/staging/usbip/usbip_common.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/staging/usbip/usbip_common.c 
> b/drivers/staging/usbip/usbip_common.c
> index 6c44fa4..faefee9 100644
> --- a/drivers/staging/usbip/usbip_common.c
> +++ b/drivers/staging/usbip/usbip_common.c
> @@ -589,7 +589,7 @@ static void correct_endian_ret_submit(struct 
> usbip_header_ret_submit *pdu,
>  be32_to_cpus(&pdu->status);
>  be32_to_cpus(&pdu->actual_length);
>  be32_to_cpus(&pdu->start_frame);
> - cpu_to_be32s(&pdu->number_of_packets);
> + be32_to_cpus(&pdu->number_of_packets);

Are you sure about this?  Why hasn't anyone else had a problem here with
this?

odd,

greg k-h